Generative AI floods the web with spam

Rapid developments in the field of artificial intelligence have made follow-up a daunting task. However, before AI can achieve certain goals, we need to understand the latest stories in the field and the important research and experiments that we have not covered.

This week, a company called SpeedyBrand quietly rose with the backing of Y Combinator. SpeedyBrand utilizes generative artificial intelligence to create SEO-optimized content. Although the company hasn't attracted a lot of money (just $2.5 million) and its relatively small customer base (about 50 brands), it got me thinking about how generative AI could begin to change the makeup of the web.

As James Vincent of The Verge pointed out in a recent article, generative AI models make it cheaper and easier to generate low-quality content. Newsguard, a company that provides news source moderation tools, has revealed many ad-supported websites that use false information created using generating artificial intelligence. This creates problems for advertisers. Newsguard's report noted that nearly 400 ads from 141 well-known brands appeared on 55 spam news sites.

However, this is not just a problem for advertisers. As Gizmodo's Kyle Barr points out, just one AI-generated article is enough to attract a lot of engagement. Even if each generated article can only generate a few dollars, this is still less than the cost of generating text, and potential advertising revenue will not be able to go to legitimate websites.

So what is the solution to this problem? Is there a workaround? These two problems keep me up at night. Barr said search engines and ad platforms have a responsibility to put in place tighter controls and punish those who use generative AI for misconduct. However, given the rapid pace of development in the field and the infinite scalability of generative AI, I'm not sure they can keep up with this progress.

Of course, spam is not a new problem, similar situations have occurred before, and the web has adapted to this phenomenon. But what's different this time is that the barrier to entry is very low, both in terms of cost and time.

James Vincent's article gives a positive view, suggesting that if the network ends up being overwhelmed by AI spam, this could prompt the development of more well-funded platforms. However, I have reservations about this. There is no doubt that we are at a critical juncture where the decisions we make about generative AI and its output will have a significant impact on network functionality for some time to come.

In addition to these issues, there are some other compelling AI stories:

OpenAI officially launches GPT-4: OpenAI this week announced the general availability of its latest text-generating model, GPT-4, available through a paid API. GPT-4 can generate text, accept images, and text input, which is an improvement over GPT-3.5, which only accepts text input and demonstrates "human level" on various professional and academic benchmarks. However, as we previously reported, it is not perfect. (ChatGPT usage has reportedly declined, but we'll have to look further.) ）

Control "superintelligent" AI: In another news from OpenAI, the company is assembling a new team, led by chief scientist and one of OpenAI's co-founders, Ilya Sutskever, to develop ways to guide and control "superintelligent" AI systems.

New York City Anti-Bias Law: After months of delays, New York City this week began enforcing a law requiring employers who use algorithms for recruitment, hiring, or promotion to submit those algorithms for independent audits and make the results public.

Valve acquiesces in AI-generated games: Valve issued a rare statement claiming that they reject games that include AI-generated assets on the Steam Games Store. The developer, known for its secrecy, says its policies are evolving, not against AI.

Humane launches Ai Pin: Startup Humane revealed details of their first product, Ai Pin. Founded by former Apple designers and engineers Imran Chaudhri and Bethany Bongiorno, the Ai Pin is a wearable device with a projection display and artificial intelligence capabilities similar to a futuristic smartphone but with a very different look.

Warning on EU AI regulation: Tech founders, CEOs, venture capitalists and industry giants across Europe signed an open letter this week warning that if the EU passes laws restricting innovation, Europe could miss out on how rapidly generative AI is rapidly changing our world, especially in the field of generative AI. However, with its development and wide application, we also face some profound reflections and challenges.

Generative AI models make it cheaper and easier to generate low-quality content, raising concerns about network composition and content quality. For example, Newsguard revealed that many ad-supported websites use generative AI to create false information, which poses a problem for advertisers. These spam sites abuse programmatic advertising or automated advertising systems to allow ads from major brands to appear in them. This not only damages the image of advertisers, but may also allow their advertising revenue to flow to illegal websites.

Another issue is that articles generated by generative AI are likely to attract a lot of engagement, although the revenue per article may be lower. However, the cost of generating these generated articles is lower than the cost of real text, so potential advertising revenue may go to low-quality content rather than legitimate websites.

Solving these problems is not easy. Search engines and ad platforms need to put in place tighter controls to punish bad actors who abuse generative AI. However, due to the rapid development of the field of artificial intelligence and the unlimited scalability of generative AI, this can be a daunting task.

Although spam is not a new problem, generative AI has made it more common and cheaper. This raises concerns about the future of the web. However, there are also some positive arguments that if the web is inundated with AI spam, this could spur the development of more well-funded platforms. However, the ultimate solution to this problem remains to be seen.

In addition to the impact of generative AI, there are other interesting AI stories. OpenAI has launched its latest text-generating model, GPT-4, which has a wider range of features and is available through a paid API. OpenAI is also developing ways to guide and control superintelligent AI systems.

On the legal and regulatory side, the City requires employers who use algorithms to recruit, hire, or promote to conduct independent audits and make the results public. In Europe, tech leaders have warned that overly restricting AI innovation could lead to missed generative AI revolutions.

In addition, artificial intelligence has shown great potential in many areas. For example, in the medical field, smart intubation systems can potentially save lives by assisting doctors with computer vision technology. AI also plays an important role in animation, archaeology, predicting natural disasters, and more.

In conclusion, the rise of generative AI has led to deep thinking about the composition of the web and the quality of content. We need to think about how to solve the problems posed by generative AI and find ways to balance development and protect the interests of users. As AI continues to evolve, we need to keep an eye on its impact and actively guide its development direction for a better future.